HacknPlan MCP Server
Enables AI assistants to manage game development projects, sprints, tasks, and design documents via the HacknPlan API.
README
HacknPlan MCP Server
A Model Context Protocol (MCP) server that provides a pure API wrapper for the HacknPlan project management API. Enables AI assistants like Claude to manage game development projects, sprints, tasks, and design documents.
✨ Key Features
- 🛡️ Deletion Recovery - 30-day recovery cache with crash resilience (v7.2.0)
- ⏱️ Automatic Time Tracking - Zero-effort time logging (v7.1.0+)
- 🚀 Context Optimization - 75-85% token reduction with slim mode + tool search
- 📦 Batch Operations - Update 1-100 items in a single call
- 🔍 Advanced Search - Multi-criteria filtering with name resolution
- 💾 Smart Caching - 80% fewer API calls with automatic 5-minute cache
Full feature list: Documentation →
🚀 Quick Start
1. Install
git clone https://github.com/yourusername/hacknplan-mcp.git
cd hacknplan-mcp
npm install
npm run build
2. Configure
claude mcp add hacknplan -s user \
-e HACKNPLAN_API_KEY=your-api-key \
-- node /path/to/hacknplan-mcp/dist/index.js
3. Use
// Create a task
await callTool('create_work_items', {
items: [{
title: 'Implement OAuth',
categoryId: 'programming', // Name resolution
importanceLevelId: 'high', // Name resolution
tagIds: ['backend', 'security'] // Auto-creates tags
}]
});
// Start working (automatic time tracking)
await callTool('start_task', { workItemId: 123 });
// Complete (auto-logs hours)
await callTool('complete_task', {
workItemId: 123,
comment: 'OAuth implementation complete'
});
📚 Documentation
→ Complete Documentation Index
Essential Guides
| Guide | Description |
|---|---|
| Quick Start | Your first API calls (5 min) |
| Work Items | Task management reference |
| Time Tracking | Automatic time logging |
| Deletion Safety | Recovery system |
| Common Patterns | Real-world recipes |
By Experience Level
Beginner: Installation → Configuration → Quick Start → Work Items
Intermediate: Time Tracking → Workflow Shortcuts → Batch Operations
Advanced: Slim Mode → Tool Search → Caching → Best Practices
🎯 Use Cases
Sprint Planning
// Create sprint
const sprint = await callTool('create_boards', {
boards: [{
name: 'Sprint 4: Authentication',
startDate: '2026-01-01T00:00:00Z',
endDate: '2026-01-14T23:59:59Z'
}]
});
// Get backlog and assign to sprint
const backlog = await callTool('get_backlog', {});
await callTool('update_work_items', {
items: backlog.items.slice(0, 10).map(item => ({
workItemId: item.workItemId,
boardId: sprint.items[0].boardId
}))
});
Automatic Time Tracking
// Start task (begins tracking)
await callTool('start_task', { workItemId: 123 });
// ... work for 2.5 hours ...
// Complete (auto-logs 2.5h)
await callTool('complete_task', {
workItemId: 123,
comment: 'Feature complete, tests passing'
});
// ✅ Automatically logged 2.5 hours
Batch Operations
// Create multiple tasks at once
await callTool('create_work_items', {
items: [
{ title: 'Design API', categoryId: 'design', estimatedCost: 4 },
{ title: 'Implement endpoints', categoryId: 'programming', estimatedCost: 8 },
{ title: 'Write tests', categoryId: 'programming', estimatedCost: 6 }
]
});
// ✅ 3 tasks created in 1 API call
🛡️ Safety Features
Deletion Recovery (v7.2.0)
All deletions automatically stored in recovery cache for 30 days:
// Delete with automatic recovery
await callTool('delete_work_items', { workItemIds: [123, 124, 125] });
// Later, recover if needed (up to 30 days)
await callTool('recover_deleted_items', { workItemIds: [123, 124, 125] });
⚡ Performance
Context Optimization
Slim Mode (75% token reduction):
HACKNPLAN_SLIM_MODE=true
HACKNPLAN_OUTPUT_VERBOSITY=slim
Tool Search (85% reduction on tool loading):
- Loads only 8 core tools initially
- Discovers 88 additional tools on-demand
- → Tool Search Guide
Smart Caching (80% fewer API calls):
- Automatic 5-minute cache on all GET requests
- → Caching Guide
📖 Complete Documentation
→ Documentation Index (docs/INDEX.md)
Core API
- Projects - Project management
- Boards & Sprints - Sprint planning
- Work Items - Tasks, bugs, features
- Comments - Markdown comments
- Milestones - Major deliverables
- Design Elements - Game design docs
Features
- Deletion Safety - Recovery system
- Automatic Time Tracking - Zero-effort logging
- Slim Mode - Context optimization
- Array-Based API - Batch operations
- Tool Search - Dynamic tool discovery
Advanced
- Workflow Shortcuts - Combined operations
- Batch Operations - Search & analytics
- Metadata & Config - Stages, tags, users
- Caching - Performance optimization
Reference
- Common Patterns - Real-world recipes
- Error Handling - Error types & recovery
- Best Practices - Optimization tips
🔧 Configuration
Environment Variables
HACKNPLAN_API_KEY=your-key # Required: API authentication
HACKNPLAN_DEFAULT_PROJECT=230954 # Optional: Default project ID
HACKNPLAN_SLIM_MODE=true # Optional: Enable 26-tool slim mode
HACKNPLAN_OUTPUT_VERBOSITY=slim # Optional: Minimal response fields
HACKNPLAN_DELETION_CACHE_SIZE=1000 # Optional: Recovery cache capacity
→ Complete Configuration Guide
🤝 Contributing
Contributions welcome! Please see:
- Issues: GitHub Issues
- Pull Requests: Follow standard GitHub PR workflow
- Documentation: Help improve docs/
📄 License
MIT License - See LICENSE for details
🔗 Links
- Documentation: docs/INDEX.md
- API Reference: API-REFERENCE.md (Complete technical reference)
- Claude.ai Guide: CLAUDE.md (AI assistant integration)
- HacknPlan API: Official Docs
- GitHub: yourusername/hacknplan-mcp
🆘 Support
Need help?
- Quick Start Guide - Get running in 5 minutes
- Common Patterns - Real-world recipes
- Error Handling - Troubleshooting guide
- GitHub Issues - Report bugs or request features
Version: 7.2.0 | Last Updated: December 21, 2025
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。